You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Usher removes seqs with >5 reversions or >20 private mutations to prevent artefact.
However, recently a lot of recombinants between DV, EG, JD, GK, FL, FU,... are popping up and sometimes these recombinants are usually removed due to having too many reversions or private mutations. The number of such removal events is quite a lot now.
And these recomb are more and more. It becomes harder and harder to separate and list every of those removed recombinants and add them back to tree.
A suggestion is that for new seqs before they are removed check for recombinants and see if it is close to any potential recomb. If it is then not remove.
Or allow manually add stationery points that prevents removal for seqs close to such points so that at least seqs for each recombinant won't be removed after the initial detection.
The text was updated successfully, but these errors were encountered:
xz-keg
changed the title
Suggestion to update removal rules to satisfy the current mass recombinant situation
Suggestion to update removal rules to adapt to the current mass recombinant situation
Sep 30, 2023
Usher removes seqs with >5 reversions or >20 private mutations to prevent artefact.
However, recently a lot of recombinants between DV, EG, JD, GK, FL, FU,... are popping up and sometimes these recombinants are usually removed due to having too many reversions or private mutations. The number of such removal events is quite a lot now.
For example
sars-cov-2-variants/lineage-proposals#846
sars-cov-2-variants/lineage-proposals#839
sars-cov-2-variants/lineage-proposals#811
sars-cov-2-variants/lineage-proposals#674
sars-cov-2-variants/lineage-proposals#879
sars-cov-2-variants/lineage-proposals#888
And these recomb are more and more. It becomes harder and harder to separate and list every of those removed recombinants and add them back to tree.
A suggestion is that for new seqs before they are removed check for recombinants and see if it is close to any potential recomb. If it is then not remove.
Or allow manually add stationery points that prevents removal for seqs close to such points so that at least seqs for each recombinant won't be removed after the initial detection.
The text was updated successfully, but these errors were encountered: